330 THE BUTTERFLIES OF NEW ENCJLAND. 



ered distinct species ; in each form, too, the sexes differ considerably in the 

 coloration of the under surface of the wings, so that the species includes 

 four sets of individuals, which may be distinguished quite as readily as a 

 great many acknowledged species of the best studied faunas ; but Mr. 

 Edwards has succeeded in rearing each from the eggs of the other, 

 and others have since done the same, leaving no doubt of their actual iden- 

 tity. Mr. Edwards alone had bred in 1882 more than twenty different 

 batches, mostly from eggs the exact parentage of which was known, and 

 raised over five hundred butterflies. In his experiments up to 1878 all the 

 eggs laid by hibernating fabricii produced umbrosa. The eggs of the sec- 

 ond brood produced 88 per cent umbrosa ; of the third brood 55 per cent 

 umbrosa ; of the fourth brood, all fabricii. Including experiments made 

 between 1878 and 1882 these proportions varied only in the second and 

 third broods, which were changed respectively to 83 and (58 per cent. This 

 is in West Virginia. In New England, where there are but two broods, 

 the difference is nearly as decisive, to judge from the few experiments known 

 to me. Here the eggs laid by the hibernating females produce nearly but 

 not quite all umbrosa ; and the eggs of the last brood almost invariably 

 produce only fabricii. ^Ir. Carl Braun for instance obtained in Bangor, 

 Me., one hundred effofs from a hibernating fabricii about the middle of June, 

 and of the residting butterflies all but two (which were fabricii 9 ) were um- 

 brosa ; while in the same year a brood raised from caterjjillars of the second 

 brood produced only fabricii, foiu- males, three females. 



Mr. F. H. Sprague's experience tells about the same story, excepting in 

 1887, when he writes me "the August brood was mixed, about evenly 

 divided between the two forms." He adds that the later ones (which he 

 looks on as a third brood) were fabricii, though an umbrosa was reared in 

 the last week in August ; so too, the same year, I bred about the middle 

 of October a single male of the form umbrosa. 



The wintering butterflies then all belong to the form fabricii, and from 

 their eggs are produced, in New England, almost exclusively the form 

 umbrosa, which in their turn yield only fabricii. The dimorphism is there- 

 fore seasonal but not exclusively so. 



A similar dimorphism affecting both the coloration and the contour of 

 the wings is found in some other species of this genus, but in none is it so 

 marked as in this, and in some it is very slight. The step from such 

 dimorphism as this species presents to the formation of distinct species 

 would seem to be very slight. 



Enemies. A large proportion of the chrysalids of this species, accord- 

 ing to Harris and Doubleday, are destroyed by Pteromalus vanessae 

 Harr. (89: 3) ; Harris obtained them from chrysalids at the end of March 

 in 1831. Oftentimes, says Mr. Bethune, "we have been disappointed in 

 our attempt to raise the butterfly of this species by this little fly. Every- 



